Research On Improvement And Application Of Butterfly Optimization Algorithm

Butterfly optimization algorithm is an optimization algorithm based on swarm intelligence proposed in 2018.Compared with classical optimization algorithms,butterfly optimization algorithm has simpler structure and fewer key parameters,and can be widely applied to practical problems.However,the optimization ability of this algorithm needs to be improved.When the problem dimension is very large,the algorithm performance will be greatly reduced,and the local search mode of the algorithm is single,so it is difficult to jump out of the local optimal value.In this paper,some effective improvement methods are proposed for the basic butterfly optimization algorithm,which can improve the optimization accuracy and the ability to jump out of the local optimal.The improved algorithm is applied to function optimization,classical engineering optimization problem and practical engineering application problem,and the application range of the algorithm is expanded.The main research contents of this paper are as follows:(1)A chaotic adaptive butterfly optimization algorithm was proposed to solve the problem that the search step size of butterfly optimization algorithm was not limited and it is difficult to jump out of the local optimal.Chaos initialization and adaptive weights were introduced to increase the diversity of the population and adjust the individual search step.Cauchy variation perturbation was performed on the globally optimal individual at the later stage of the algorithm.Through the comparison test of 12 benchmark test functions and the optimal cost design experiment of welded beam,it was proved that the improved algorithm has strong optimization ability and has certain advantages in the application field.(2)In order to improve the search accuracy and efficiency of the butterfly optimization algorithm,and expand the application field of the algorithm,this paper proposed a golden sine weighted butterfly optimization algorithm.Firstly,the golden sine operator was introduced to reduce the solution space of the algorithm in the search process.Secondly,the adaptive weight was used to adjust the step size and direction of the population individual movement.By optimizing the hydrogeological parameters,the experimental results showed that the improved algorithm can effectively reduce the pumping depth reduction error of single observation hole and multiple observation holes to 0.006195 and 0.0030912 respectively,and improved the calculation performance of Theis formula,which provided a new method for subsequent pumping experiments.(3)Aiming at the shortcomings of low search efficiency and low convergence accuracy of the algorithm,this paper proposed an improved butterfly optimization algorithm with multiple strategies.Firstly,the finder and entrant strategy of sparrow search algorithm was introduced,and then the adaptive weight coefficient was used to improve the optimization accuracy.Finally,the global optimal individual was perturbed by Cauchy dimensional mutation to increase the population diversity in the later stage of the algorithm.The improved algorithm was applied to the node coverage optimization of wireless sensor network.The results showed that the optimized node distribution was more uniform and the coverage rate was up to99.41%,which can effectively reduce the cost of wireless sensor network configuration.